The Invisible Woman. Joanne Belknap
This chart summarizes three theories that evolved around the same time, primarily as they started. Fortunately, there have been improvements; for example, LCT now usually includes girls/women and increasingly included childhood abuse, and CVT is more likely to include self-report offending behaviors. Chapter 14 addresses the way these theories are more commonly combined or paired in more recent research.
Data collected consistent with LCT is prospective, collecting data over time on the same individuals. CVT is also considered prospective, typically using official data on the same individuals over time, such as childhood victimizations reported to hospitals and the police and adolescent and adult arrest records, although increasingly, CVT studies also use self-report data (e.g., A. A. Fagan, 2005; Wright & Fagan, 2013). Conversely, PT research is retrospective, asking individuals about their past experiences at one point in time. The advantages of prospective data collection are improved accuracy of the temporal ordering of life events (e.g., “Did the child run away or start using drugs before or after being sexually abuse?”) and the inclusion of people who will not offend so that resiliency can be studied and better understood (e.g., “What was related to not offending?”). PT is far less expensive than CVT or LCT because the participants are surveyed or interviewed at only one point in time about their lives. But PT studies usually are conducted solely on incarcerated offenders, so they offer limited resiliency analysis other than looking at desistance over time (which is still important, but CVT and LCT can do this, as well). An advantage of PT is that incidents of abuse and neglect will likely be more accurate and frequently reported than with LCT or CVT. In LCT, children who are asked about these incidents may feel unsafe or uncomfortable reporting them to a researcher, and adults (e.g., parents, guardians, teachers) who are asked about children’s victimizations may not know about them or may even be the perpetrators of them. PT data on child maltreatment is likely more accurate than CVT data, which has historically relied on official reports (and most child abuse is not officially reported). Of course, whether prospective or retrospective, a survivor of trauma can also completely repress such events (even officially reported ones), so they may never be known (unless there is official documentation).
Cycle of Violence Theory (CVT)
Widom developed CVT in 1989 to assess the “scholarly claims that abused children become abusers, and victims of violence become violent victimizers” (Widom, 1989b, p. 355). Her prospective data included 908 abused/neglected youths (youths for which there was a court-validated physical, sexual, and/or neglect victimization before age 12) with 667 controls (youths with none of these victimizations prior to age 12, matched by gender, age, race, and hospital of birth) and their official juvenile and adult arrests. As expected with CVT, compared to controls, abused/neglected youths had “more arrests as a juvenile (26 vs. 17%), more arrests as an adult (29 vs. 21%), and more arrests for any violent offense (11 vs. 8%)” (Widom (1989b, p. 244). Widom (1989a, 1989b) found the relationship between abuse/neglect was gendered, in that abused/neglected boys were even more likely than abused/neglected girls to become adult criminals and to become violent criminals. Although abused/neglected boys were no more likely than the control boys to be arrested for violent offenses as juveniles, they were more likely to have adult arrests, while abused/neglected girls were no more likely than control girls to have adult violent offense arrests, “but there was a trend toward increased violence [arrests] as juveniles” (Rivera & Widom, 1990, p. 24).
Widom (1995) found that although childhood sexual abuse victimization placed an individual at increased risk of future arrests, these childhood abuse survivors were no more likely to be arrested later in life than were the individuals who experienced no (officially reported) sexual abuse but experienced (officially reported) physical abuse and/or neglect. However, “sexual abuse plus” victims (i.e., those who experienced sexual abuse plus physical abuse and/or neglect as a child) were at the greatest risk of being arrested for running away. Moreover, victims of childhood sexual abuse were far more likely to be arrested for prostitution than their counterparts who were not sexually abused. (However, arrests for running away did not significantly predispose youths to prostitution arrests.) Analyses of the individual effects of three types of maltreatment showed that “any sexual abuse” had the greatest impact on the likelihood of being arrested for prostitution, and “any nonsexual physical abuse” had the greatest impact on the likelihood of being arrested for rape or sodomy. Unfortunately, these statistics lump females and males together, so we do not know how these dynamics may vary by gender. Updating the data, Widom and Maxfield (2001) found abused or neglected girls were 74% more likely than non-abused/non-neglected girls to be arrested for property, alcohol, drug and misdemeanor crimes, and over twice as likely to be arrested for violent crimes.
Widom’s contributions are huge, but it is important to remember that she relied on court-documented cases of child abuse and neglect; thus, her studies do not include the many cases that do not reach the courts and likely disproportionately reflect the most extreme child abuse cases (Widom, 1995). Additionally, River and Widom (1990, p. 25) found that childhood abuse/neglect significantly increased the likelihood of Black but not white youths’ adult violent arrests, but relying on arrests does not take into account the differential (racist) processing of Black compared to white individuals (see Chapter 6). Alternatively, two extensive tests of CVT, using longitudinal data self-report (not official) victimization and offending data, and controlling for many variables, found that among youths surviving childhood physical abuse, boys/men are still more likely than girls/women to subsequently commit violence (A. A. Fagan, 2005; Wright & Fagan, 2013). Another prospective study consistent with CVT (but like Widom, using official victimization and offending data) included all cases of girl victims of child sexual abuse who went to an emergency room in the 1970s in a major northeastern city, matched them with girls with no such histories, and then compared their offense histories (Siegel & Williams, 2003). Consistent with PT, those sexually abused as girls were more likely to be arrested for running away (oddly, though, this was more often for cases where the girl was sexually abused by a stranger than by a family member), drug offenses, property offenses, and prostitution (Siegel & Williams, 2003).
Widom, Fisher, Nagin, and Piquero’s (2018) test of CVT followed Widom’s original sample into their 50s, providing a longer-range impact analysis of childhood abuse and neglect. They found that childhood maltreatments and being male increased the likelihood of offending even into the individuals’ 50s. Distinguishing between non-offenders, low-level chronic offenders, and mid-level chronic offenders, they found maltreated females were far more prevalent among the mid-level chronic offenders than among the low-level chronic offenders (and, of course, among the non-offender group), such that abuse/neglect increased both the longevity and frequency of offending among women (p. 841). Importantly, they also found that increased women’s desistance from offending (relative to men) may be due to their higher rates of death—not actual desistance—stressing the need for research addressing the effect of child maltreatment on mortality and how it may be gendered.
Life Course Theory (LCT)
Sampson and Laub’s (1993) construction of LCT drew significantly from Hirschi’s (1969) social control theory (SCT), Gottfredson and Hirschi’s (1990) general theory of crime (GTC), and C. Taylor’s (2001) self-control theory. LCT theorizes that various life events, particularly those during childhood and adolescence, affect one’s risk of offending behavior. Thus, various developmental stages are “age specific,” making offending behavior associated with age (see Loeber, 1996). Indeed, adolescence is identified as a particularly at-risk time given the angst of puberty, the stress of schools (including changing schools), and peer pressure.
In this sense, crime is viewed as a network of various causal factors. One aspect of LCT is that independent variables become dependent variables over time. For example, delinquency decreases one’s chance of doing well in school, which in turn becomes a predictor for (re)turning to crime. Some of the key variables assessed in much of the life course research include antisocial behavior, intelligence, and income levels, as well as general criminal, delinquent, and deviant behaviors. Thus, a focus of some life course research